This application, submitted in response to PAR-03-078, From Intervention Development to Services: Exploratory Research Grants (R34), seeks funding for a 3 year project to adapt and field test two commonly-used and efficacious community-level HIV prevention interventions, Integrated Outreach and the Popular Opinion Leader (POL) Model, for use in online environments such as "chat rooms." It is becoming increasingly common for high-risk individuals to meet sexual partners online. In multiple studies with diverse populations, researchers have documented that persons who seek sexual partners online show a pattern of increased sexual risk behaviors. The Internet also has potential utility to reach populations at risk for HIV. We believe that community-level interventions (Integrated Outreach, POL) which intervene with populations in their natural environment during their day-to-day interactions show particular promise for adaptation for online use. To examine this, we will conduct 3 phases of research: (1) initial interviews and focus groups with persons using chat rooms and with providers who conduct online interventions (2) using information obtained during the formative stage, we will develop treatment protocols and standardized intervention manuals building upon our preliminary conceptualization of these interventions as adapted for the Internet, (3) we will conduct pilot studies and a small-scale test-of-concept field test to examine the potential efficacy of the interventions to alter community norms, attitudes, and HIV risk behavior. The field test will use a randomized design with 4 chat sites randomly assigned to 1 of 3 conditions: (a) Integrated Outreach, (b) POL, and (c) assessment-only control. Consistent with the PA, methods used will provide preliminary data concerning the feasibility, acceptability, and practicality of online community-level interventions. The proposed methods will also permit us to collect data needed to provide initial evidence of efficacy needed for a larger-scale trial, including estimates of outcome parameters such as intervention effect sizes.